[bug report] blktests nvme/fc nvme/057 triggered kernel panic

[Daniel Wagner]

On Wed, Aug 20, 2025 at 10:36:15AM +0200, Daniel Wagner wrote:
> On Wed, Aug 20, 2025 at 04:20:40PM +0800, Yi Zhang wrote:
> > My recent blktests testing triggered the kernel panic issue with
> > kernel 6.17.0-rc2.
> > Please help check it and let me know if you need any info/testing for
> > it, thanks.
> > 
> > [1] reproducer
> > stress blktests nvme/fc nvme/057
> 
> I will look into it. I've got a similar bug report from our internal QA
> which I have debugged but not fixed yet.
> 
> The async handling code is not able to handle more than one in-flight
> command (__nvmet_fc_send_ls_req). I was able to trigger the bug by
> running nvme/044 for the FC transport in a tight loop.
> 
> This might be related here, as it seems to happen when disconnecting
> from the target, similar to the stacktrace I see with nvme/044.

It turns out this bug is a different one that I fixed in

https://lore.kernel.org/linux-nvme/20250821-fix-nvmet-fc-v1-1-3349da4f416e@kernel.org/

I'll try to figure out what's happening here.